Sapia.ai, creator of the world’s first AI Smart Interviewer, has released a communication score update that will enable its Chat Interview tool to score candidate language skills with high accuracy.
With this innovation, Sapia can accurately score language fluency and complexity from candidate job interview responses, giving employers more data on the ability and communication skills of applicants – information which can be used to attain better quality of hire, as well as other performance indicators.
The update further cements Sapia’s AI system – which uses Natural Language Processing (NLP) and machine learning – as the only recruitment tool of its kind to leverage both text chat responses and communication scoring as a solution for multimeasure candidate screening.
Sapia.ai founder and CEO Barb Hyman said that the innovation will give employers and talent acquisition managers more certainty around hire quality and soft skills.
“We know that most, if not all, roles require good communication skills,”Hyman said. “Now, alongside behavioral competencies and personality traits, Sapia.ai customers can gauge communication skills and use all of these combined data to make even better informed decisions about who they hire.”
“Sapia’s text-based approach to candidate interviewing and assessment has transformed hiring for our customers, affecting considerable savings to the time and costs associated with recruitment. By using text, customers no longer need to screen resumes or review face-to-face interview correspondence, instead relying on objective, unbiased data. Most of all candidates love the familiar asynchronous chat interaction”, Hyman said.
The communication score update was developed through rigorous academic research, testing, and validation. Dr Madhura Jayaratne, Lead Data Scientist at Sapia.ai who led the research, said that bias testing was an important part of the development process.
“We acknowledge that our interview tool is used by many different kinds of people whose written communication skills cover a wide spectrum. It is critical that our algorithm does not unfairly disadvantage candidates who belong to different demographic groups and various job families. Our candidate base of over two million candidates from around the globe allow us to test and benchmark across different groups,” Dr Jayaratne said.
The communication score algorithm is based on several dimensions, including the length of responses, readability, word usage, word choice, development and organization of ideas, and others. Customers can find where each candidate stands in the communications skills benchmark within the Talent Insights report generated at the end of the Chat Interview.